# atheris

Sets up and runs Atheris, the coverage-guided Python fuzzer built on libFuzzer. Covers TestOneInput harnesses, FuzzedDataProvider, instrumenting both pure Python and native C extensions, and running under AddressSanitizer. Use when fuzzing a Python package, hunting memory corruption in a Python C extension, or choosing between Atheris and Hypothesis for a Python target.

- **Kind:** skill
- **Source:** https://github.com/trailofbits/skills
- **Page:** https://forefy.com/skills/b53f18dc-0abc-4e49-bd73-bd9a99c6d98d
- **API (JSON + files):** https://forefy.com/api/asr/b53f18dc-0abc-4e49-bd73-bd9a99c6d98d

---

## SKILL.md

---
name: atheris
type: fuzzer
description: "Sets up and runs Atheris, the coverage-guided Python fuzzer built on libFuzzer. Covers TestOneInput harnesses, FuzzedDataProvider, instrumenting both pure Python and native C extensions, and running under AddressSanitizer. Use when fuzzing a Python package, hunting memory corruption in a Python C extension, or choosing between Atheris and Hypothesis for a Python target."
---

# Atheris

Atheris is a coverage-guided Python fuzzer built on libFuzzer. It enables fuzzing of both pure Python code and Python C extensions with integrated AddressSanitizer support for detecting memory corruption issues.

## When to Use

| Fuzzer | Best For | Complexity |
|--------|----------|------------|
| Atheris | Python code and C extensions | Low-Medium |
| Hypothesis | Property-based testing | Low |
| python-afl | AFL-style fuzzing | Medium |

**Choose Atheris when:**
- Fuzzing pure Python code with coverage guidance
- Testing Python C extensions for memory corruption
- Integration with libFuzzer ecosystem is desired
- AddressSanitizer support is needed

## Quick Start

```python
import sys
import atheris

@atheris.instrument_func
def TestOneInput(data: bytes):
    if len(data) == 4:
        if data[0] == 0x46:  # "F"
            if data[1] == 0x55:  # "U"
                if data[2] == 0x5A:  # "Z"
                    if data[3] == 0x5A:  # "Z"
                        raise RuntimeError("You caught me")

def main():
    atheris.Setup(sys.argv, TestOneInput)
    atheris.Fuzz()

if __name__ == "__main__":
    main()
```

Run:
```bash
uv run python fuzz.py
```

## Installation

Atheris supports 32-bit and 64-bit Linux, and macOS. We recommend fuzzing on Linux because it's simpler to manage and often faster.

### Prerequisites

- Python 3.7 or later
- Recent version of clang (preferably [latest release](https://github.com/llvm/llvm-project/releases))
- For Docker users: [Docker Desktop](https://www.docker.com/products/docker-desktop/)

### Linux/macOS

```bash
uv init --bare   # once, if the harness directory is not yet a uv project
uv add atheris
```

### Docker Environment (Recommended)

For a fully operational Linux environment with all dependencies configured:

```dockerfile
# https://hub.docker.com/_/python
ARG PYTHON_VERSION=3.11

FROM python:$PYTHON_VERSION-slim-bookworm

RUN python --version

RUN apt update && apt install -y \
    ca-certificates \
    wget \
    && rm -rf /var/lib/apt/lists/*

# LLVM builds version 15-19 for Debian 12 (Bookworm)
# https://apt.llvm.org/bookworm/dists/
ARG LLVM_VERSION=19

RUN echo "deb http://apt.llvm.org/bookworm/ llvm-toolchain-bookworm-$LLVM_VERSION main" > /etc/apt/sources.list.d/llvm.list
RUN echo "deb-src http://apt.llvm.org/bookworm/ llvm-toolchain-bookworm-$LLVM_VERSION main" >> /etc/apt/sources.list.d/llvm.list
RUN wget -qO- https://apt.llvm.org/llvm-snapshot.gpg.key > /etc/apt/trusted.gpg.d/apt.llvm.org.asc

RUN apt update && apt install -y \
    build-essential \
    clang-$LLVM_VERSION \
    && rm -rf /var/lib/apt/lists/*

ENV APP_DIR "/app"
RUN mkdir $APP_DIR
WORKDIR $APP_DIR

ENV VIRTUAL_ENV "/opt/venv"
RUN python -m venv $VIRTUAL_ENV
ENV PATH "$VIRTUAL_ENV/bin:$PATH"

# https://github.com/google/atheris/blob/master/native_extension_fuzzing.md#step-1-compiling-your-extension
ENV CC="clang-$LLVM_VERSION"
ENV CFLAGS "-fsanitize=address,fuzzer-no-link"
ENV CXX="clang++-$LLVM_VERSION"
ENV CXXFLAGS "-fsanitize=address,fuzzer-no-link"
ENV LDSHARED="clang-$LLVM_VERSION -shared"
ENV LDSHAREDXX="clang++-$LLVM_VERSION -shared"
ENV ASAN_SYMBOLIZER_PATH="/usr/bin/llvm-symbolizer-$LLVM_VERSION"

# Allow Atheris to find fuzzer sanitizer shared libs
# https://github.com/google/atheris#building-from-source
RUN LIBFUZZER_LIB=$($CC -print-file-name=libclang_rt.fuzzer_no_main-$(uname -m).a) \
    python -m pip install --no-binary atheris atheris

# https://github.com/google/atheris/blob/master/native_extension_fuzzing.md#option-a-sanitizerlibfuzzer-preloads
ENV LD_PRELOAD "$VIRTUAL_ENV/lib/python3.11/site-packages/asan_with_fuzzer.so"

# 1. Skip memory allocation failures for now, they are common, and low impact (DoS)
# 2. https://github.com/google/atheris/blob/master/native_extension_fuzzing.md#leak-detection
ENV ASAN_OPTIONS "allocator_may_return_null=1,detect_leaks=0"

CMD ["/bin/bash"]
```

Build and run:
```bash
docker build -t atheris .
docker run -it atheris
```

### Verification

```bash
python -c "import atheris; print(atheris.__version__)"
```

## Writing a Harness

### Harness Structure for Pure Python

```python
import sys
import atheris

@atheris.instrument_func
def TestOneInput(data: bytes):
    """
    Fuzzing entry point. Called with random byte sequences.

    Args:
        data: Random bytes generated by the fuzzer
    """
    # Add input validation if needed
    if len(data) < 1:
        return

    # Call your target function
    try:
        your_target_function(data)
    except ValueError:
        # Expected exceptions should be caught
        pass
    # Let unexpected exceptions crash (that's what we're looking for!)

def main():
    atheris.Setup(sys.argv, TestOneInput)
    atheris.Fuzz()

if __name__ == "__main__":
    main()
```

### Structured Input with FuzzedDataProvider

A target taking several typed arguments wastes most of the fuzzer's inputs if the harness
slices `data` by hand, because every mutation shifts the byte offsets of everything after it.
`atheris.FuzzedDataProvider` splits one `bytes` input into typed values instead:

```python
fdp = atheris.FuzzedDataProvider(data)
name = fdp.ConsumeUnicodeNoSurrogates(fdp.ConsumeIntInRange(0, 64))
strict = fdp.ConsumeBool()
```

See [structured-input.md](structured-input.md) for the full method reference, the fixed-draw-
order rule, and what each method returns once the buffer runs dry.

### Harness Rules

| Do | Don't |
|----|-------|
| Use `@atheris.instrument_func` for coverage | Forget to instrument target code |
| Catch expected exceptions | Catch all exceptions indiscriminately |
| Use `atheris.instrument_imports()` for libraries | Import modules after `atheris.Setup()` |
| Keep harness deterministic | Use randomness or time-based behavior |

> **See Also:** For detailed harness writing techniques, patterns for handling complex inputs,
> and advanced strategies, see the **fuzz-harness-writing** technique skill.

## Fuzzing Pure Python Code

For fuzzing broader parts of an application or library, use instrumentation functions:

```python
import atheris
with atheris.instrument_imports():
    import your_module
    from another_module import target_function

def TestOneInput(data: bytes):
    target_function(data)

atheris.Setup(sys.argv, TestOneInput)
atheris.Fuzz()
```

**Instrumentation Options:**
- `atheris.instrument_func` - Decorator for single function instrumentation
- `atheris.instrument_imports()` - Context manager for instrumenting all imported modules
- `atheris.instrument_all()` - Instrument all Python code system-wide

## Fuzzing Python C Extensions

Python C extensions require compilation with specific flags for instrumentation and sanitizer support.

### Environment Configuration

If using the provided Dockerfile, these are already configured. For local setup:

```bash
export CC="clang"
export CFLAGS="-fsanitize=address,fuzzer-no-link"
export CXX="clang++"
export CXXFLAGS="-fsanitize=address,fuzzer-no-link"
export LDSHARED="clang -shared"
```

### Example: Fuzzing cbor2

Install the extension from source:
```bash
CBOR2_BUILD_C_EXTENSION=1 uv add --no-binary-package cbor2 'cbor2==5.6.4'
```

The `--no-binary-package` flag ensures the C extension is compiled locally with
instrumentation rather than pulled as a prebuilt wheel. Persist that choice with
`no-binary-package = ["cbor2"]` under `[tool.uv]` in `pyproject.toml`, or a later
`uv sync` can silently swap in an uninstrumented wheel.

Create `cbor2-fuzz.py`:
```python
import sys
import atheris

# _cbor2 ensures the C library is imported
from _cbor2 import loads

def TestOneInput(data: bytes):
    try:
        loads(data)
    except Exception:
        # We're searching for memory corruption, not Python exceptions
        pass

def main():
    atheris.Setup(sys.argv, TestOneInput)
    atheris.Fuzz()

if __name__ == "__main__":
    main()
```

Run:
```bash
uv run python cbor2-fuzz.py
```

> **Important:** When running locally (not in Docker), you must [set `LD_PRELOAD` manually](https://github.com/google/atheris/blob/master/native_extension_fuzzing.md#option-a-sanitizerlibfuzzer-preloads).

## Corpus Management

### Creating Initial Corpus

```bash
mkdir corpus
# Add seed inputs
echo "test data" > corpus/seed1
echo '{"key": "value"}' > corpus/seed2
```

Run with corpus:
```bash
uv run python fuzz.py corpus/
```

### Corpus Minimization

Atheris inherits corpus minimization from libFuzzer:
```bash
uv run python fuzz.py -merge=1 new_corpus/ old_corpus/
```

> **See Also:** For corpus creation strategies, dictionaries, and seed selection,
> see the **fuzzing-corpus** technique skill.

## Running Campaigns

### Basic Run

```bash
uv run python fuzz.py
```

### With Corpus Directory

```bash
uv run python fuzz.py corpus/
```

### Common Options

```bash
# Run for 10 minutes
uv run python fuzz.py -max_total_time=600

# Limit input size
uv run python fuzz.py -max_len=1024

# Run with multiple workers
uv run python fuzz.py -workers=4 -jobs=4
```

### Interpreting Output

| Output | Meaning |
|--------|---------|
| `NEW    cov: X` | Found new coverage, corpus expanded |
| `pulse  cov: X` | Periodic status update |
| `exec/s: X` | Executions per second (throughput) |
| `corp: X/Yb` | Corpus size: X inputs, Y bytes total |
| `ERROR: libFuzzer` | Crash detected |

## Sanitizer Integration

### AddressSanitizer (ASan)

AddressSanitizer is automatically integrated when using the provided Docker environment or when compiling with appropriate flags.

For local setup:
```bash
export CFLAGS="-fsanitize=address,fuzzer-no-link"
export CXXFLAGS="-fsanitize=address,fuzzer-no-link"
```

Configure ASan behavior:
```bash
export ASAN_OPTIONS="allocator_may_return_null=1,detect_leaks=0"
```

### LD_PRELOAD Configuration

For native extension fuzzing:
```bash
export LD_PRELOAD="$(python -c 'import atheris; import os; print(os.path.join(os.path.dirname(atheris.__file__), "asan_with_fuzzer.so"))')"
```

> **See Also:** For detailed sanitizer configuration, common issues, and advanced flags,
> see the **address-sanitizer** and **undefined-behavior-sanitizer** technique skills.

### Common Sanitizer Issues

| Issue | Solution |
|-------|----------|
| `LD_PRELOAD` not set | Export `LD_PRELOAD` to point to `asan_with_fuzzer.so` |
| Memory allocation failures | Set `ASAN_OPTIONS=allocator_may_return_null=1` |
| Leak detection noise | Set `ASAN_OPTIONS=detect_leaks=0` |
| Missing symbolizer | Set `ASAN_SYMBOLIZER_PATH` to `llvm-symbolizer` |

## Advanced Usage

### Tips and Tricks

| Tip | Why It Helps |
|-----|--------------|
| Use `atheris.instrument_imports()` early | Ensures all imports are instrumented for coverage |
| Start with small `max_len` | Faster initial fuzzing, gradually increase |
| Use dictionaries for structured formats | Helps fuzzer understand format tokens |
| Run multiple parallel instances | Better coverage exploration |

### Custom Instrumentation

Fine-tune what gets instrumented:
```python
import atheris

# Instrument only specific modules
with atheris.instrument_imports():
    import target_module
# Don't instrument test harness code

def TestOneInput(data: bytes):
    target_module.parse(data)
```

### Performance Tuning

| Setting | Impact |
|---------|--------|
| `-max_len=N` | Smaller values = faster execution |
| `-workers=N -jobs=N` | Parallel fuzzing for faster coverage |
| `ASAN_OPTIONS=fast_unwind_on_malloc=0` | Better stack traces, slower execution |

### UndefinedBehaviorSanitizer (UBSan)

Add UBSan to catch additional bugs:
```bash
export CFLAGS="-fsanitize=address,undefined,fuzzer-no-link"
export CXXFLAGS="-fsanitize=address,undefined,fuzzer-no-link"
```

Note: Modify flags in Dockerfile if using containerized setup.

## Real-World Examples

Two complete harnesses — a pure-Python parser and an HTTP response parser — are in
[examples.md](examples.md).

## Troubleshooting

| Problem | Cause | Solution |
|---------|-------|----------|
| No coverage increase | Poor seed corpus or target not instrumented | Add better seeds, verify `instrument_imports()` |
| Slow execution | ASan overhead or large inputs | Reduce `max_len`, use `ASAN_OPTIONS=fast_unwind_on_malloc=1` |
| Import errors | Modules imported before instrumentation | Move imports inside `instrument_imports()` context |
| Segfault without ASan output | Missing `LD_PRELOAD` | Set `LD_PRELOAD` to `asan_with_fuzzer.so` path |
| Build failures | Wrong compiler or missing flags | Verify `CC`, `CFLAGS`, and clang version |

## Related Skills

### Technique Skills

| Skill | Use Case |
|-------|----------|
| **fuzz-harness-writing** | Detailed guidance on writing effective harnesses |
| **address-sanitizer** | Memory error detection during fuzzing |
| **undefined-behavior-sanitizer** | Catching undefined behavior in C extensions |
| **coverage-analysis** | Measuring and improving code coverage |
| **fuzzing-corpus** | Building and managing seed corpora |

### Related Fuzzers

| Skill | When to Consider |
|-------|------------------|
| **hypothesis** | Property-based testing with type-aware generation |
| **python-afl** | AFL-style fuzzing for Python when Atheris isn't available |

## Resources

### Key External Resources

**[Atheris GitHub Repository](https://github.com/google/atheris)**
Official repository with installation instructions, examples, and documentation for fuzzing both pure Python and native extensions.

**[Native Extension Fuzzing Guide](https://github.com/google/atheris/blob/master/native_extension_fuzzing.md)**
Comprehensive guide covering compilation flags, LD_PRELOAD setup, sanitizer configuration, and troubleshooting for Python C extensions.

**[Continuously Fuzzing Python C Extensions](https://blog.trailofbits.com/2024/02/23/continuously-fuzzing-python-c-extensions/)**
Trail of Bits blog post covering CI/CD integration, ClusterFuzzLite setup, and real-world examples of fuzzing Python C extensions in continuous integration pipelines.

**[ClusterFuzzLite Python Integration](https://google.github.io/clusterfuzzlite/build-integration/python-lang/)**
Guide for integrating Atheris fuzzing into CI/CD pipelines using ClusterFuzzLite for automated continuous fuzzing.

### Video Resources

Videos and tutorials are available in the main Atheris documentation and libFuzzer resources.

## agents

```

```

## agents/openai.yaml

```yaml
interface:
  icon_small: "assets/trail-of-bits-mark.svg"
  icon_large: "assets/trail-of-bits-mark.svg"
  brand_color: "#D83A34"
```

## assets

```

```

## assets/trail-of-bits-mark.svg

```

```

## examples.md

# Atheris Examples

Two complete harnesses, each runnable as written.

## Example: Pure Python Parser

```python
import sys
import atheris
import json

@atheris.instrument_func
def TestOneInput(data: bytes):
    try:
        # Fuzz Python's JSON parser
        json.loads(data.decode('utf-8', errors='ignore'))
    except (ValueError, UnicodeDecodeError):
        pass

def main():
    atheris.Setup(sys.argv, TestOneInput)
    atheris.Fuzz()

if __name__ == "__main__":
    main()
```

## Example: HTTP Request Parsing

```python
import sys
import atheris

with atheris.instrument_imports():
    from urllib3 import HTTPResponse
    from io import BytesIO

def TestOneInput(data: bytes):
    try:
        # Fuzz HTTP response parsing
        fake_response = HTTPResponse(
            body=BytesIO(data),
            headers={},
            preload_content=False
        )
        fake_response.read()
    except Exception:
        pass

def main():
    atheris.Setup(sys.argv, TestOneInput)
    atheris.Fuzz()

if __name__ == "__main__":
    main()
```

Both catch `Exception` broadly to get a campaign started. Narrow that to the exceptions the
target is documented to raise before you trust the results — a bare `except Exception` also
swallows the bugs you are fuzzing for.

## structured-input.md

# Structured Input with FuzzedDataProvider

A target that takes several typed arguments — a string, a length, a flag — wastes most of
the fuzzer's inputs if the harness slices `data` by hand, because every mutation shifts the
byte offsets of everything after it. `atheris.FuzzedDataProvider` splits one `bytes` input
into typed values while keeping each draw stable under mutation.

## Basic Usage

```python
@atheris.instrument_func
def TestOneInput(data: bytes):
    fdp = atheris.FuzzedDataProvider(data)
    name = fdp.ConsumeUnicodeNoSurrogates(fdp.ConsumeIntInRange(0, 64))
    count = fdp.ConsumeIntInRange(1, 1000)
    strict = fdp.ConsumeBool()
    your_target_function(name, count, strict=strict)
```

Draw in a fixed order. Each call consumes from where the last one left off, so inserting or
reordering a call reinterprets every byte after it and devalues the corpus already built.

Ask for a size before the content it bounds, as above: an unbounded string lets the fuzzer
spend the whole buffer on one field and starve every draw after it.

## Method Reference

| Need | Call |
|------|------|
| Raw bytes | `ConsumeBytes(count)` |
| Text | `ConsumeUnicodeNoSurrogates(count)` |
| Text, including unpaired surrogates | `ConsumeUnicode(count)` |
| Bounded integer | `ConsumeIntInRange(min, max)` |
| Sized integer | `ConsumeInt(size)` (signed), `ConsumeUInt(size)` |
| Float | `ConsumeFloat()`, `ConsumeRegularFloat()` (no `NaN`/`Inf`), `ConsumeProbability()` |
| Flag | `ConsumeBool()` |
| Choice from a fixed set | `PickValueInList(values)` |
| Everything not yet consumed | `ConsumeBytes(fdp.remaining_bytes())` |

`remaining_bytes()` is an accessor, not a draw — it returns the count of unconsumed bytes and
consumes nothing. It is the one method here that is not named `Consume*`, and the only one
whose return value is a length rather than a value. Pass it to `ConsumeBytes` to drain the
buffer; using it directly hands your target an `int` where it expects `bytes`.

Prefer `ConsumeUnicodeNoSurrogates` unless you are specifically testing surrogate handling.
`ConsumeUnicode` may emit unpaired surrogates (U+D800–U+DFFF), which are legal in a Python
`str` but raise `UnicodeEncodeError` the moment the target encodes them — so a target that
encodes anywhere reports a crash on its own input handling rather than on your target's logic.

List variants take the element count first:

| Call | Produces |
|------|----------|
| `ConsumeIntList(count, bytes)` | `count` integers of `bytes` size each |
| `ConsumeIntListInRange(count, min, max)` | `count` integers in `[min, max]` |
| `ConsumeFloatList(count)` | `count` arbitrary floats, `NaN` and `Inf` included |
| `ConsumeRegularFloatList(count)` | `count` floats, never `NaN` or `Inf` |
| `ConsumeProbabilityList(count)` | `count` floats in `[0, 1]` |
| `ConsumeFloatListInRange(count, min, max)` | `count` floats in `[min, max]` |

## Running Out of Input

Every method degrades rather than raising when the buffer empties: consumers return empty
values, and `ConsumeIntInRange` returns `min`. A harness that draws more than the fuzzer
supplies will not error — it will quietly test the same degenerate case over and over, which
looks like a healthy campaign that has stopped finding anything. Check `remaining_bytes()`
and return early if your harness needs a minimum amount of input.